VIRTUAL GPU SOFTWARE R390 FOR CITRIX XENSERVER

VIRTUAL GPU SOFTWARE R390 FOR CITRIX XENSERVER RN-06927-001 _v6.0 Revision 02 April 2018 Release Notes

TABLE OF CONTENTS Chapter 1. Release Notes... 1 1.1. Updates in Release 6.0... 1 Chapter 2. Validated Platforms...3 2.1. Supported NVIDIA GPUs and Validated Server Platforms... 3 2.2. Hypervisor Software Releases... 4 2.3. Guest OS Support... 5 2.3.1. Windows Guest OS Support... 5 2.3.2. Linux Guest OS Support... 6 2.4. XenMotion with vgpu Support...7 Chapter 3. Known Product Limitations... 8 3.1. vgpu profiles with 512 Mbytes or less of frame buffer support only 1 virtual display head on Windows 10... 8 3.2. NVENC requires at least 1 Gbyte of frame buffer... 9 3.3. VM running older NVIDIA vgpu drivers fails to initialize vgpu when booted...9 3.4. Virtual GPU fails to start if ECC is enabled... 10 3.5. Single vgpu benchmark scores are lower than pass-through GPU... 12 3.6. nvidia-smi fails to operate when all GPUs are assigned to GPU pass-through mode... 13 3.7. Windows Aero is disabled on XenDesktop session using 3 or 4 monitors in 2560 1600 resolution... 14 3.8. VMs configured with large memory fail to initialize vgpu when booted... 14 3.9. vgpu host driver RPM upgrade fails... 15 Chapter 4. Resolved Issues... 16 Chapter 5. NVIDIA Software Security Updates... 17 Chapter 6. Known Issues... 18 6.1. Migrating a VM Configured with vgpu Fails... 18 6.2. vgpu guest VM driver not properly loaded on servers with more than 512 GB of system memory...18 6.3. Issues in remote desktop sessions if a license is acquired after a session is started...19 6.4. The license expires prematurely in Linux guest VMs... 19 6.5. Multiple display heads are not detected by Ubuntu 14.04 guest VMs...20 6.6. CILP is not working in Windows VMs...21 6.7. Luxmark causes a segmentation fault on an unlicensed Linux client... 21 6.8. On GPUs based on the Pascal architecture, Ubuntu 16.04 VMs run slowly after acquiring a license... 22 6.9. Resolution is not updated after a VM acquires a license and is restarted...22 6.10. On Tesla P40, P6, and P4 GPUs, the default ECC setting prevents NVIDIA vgpu from starting... 23 6.11. NVIDIA vgpu encoder and process utilization counters don't work with Windows Performance Counters... 23 RN-06927-001 _v6.0 Revision 02 ii

6.12. A segmentation fault in DBus code causes nvidia-gridd to exit on Red Hat Enterprise Linux and CentOS... 24 6.13. No Manage License option available in NVIDIA X Server Settings by default... 25 6.14. The nvidia-gridd service fails because the required configuration is not provided... 25 6.15. The Apply button is disabled after change to unlicensed mode... 26 6.16. Licenses remain checked out when VMs are forcibly powered off...27 6.17. Memory exhaustion can occur with vgpu profiles that have 512 Mbytes or less of frame buffer... 28 6.18. VM bug checks after the guest VM driver for Windows 10 RS2 is installed...29 6.19. On XenServer 7.0, VMs unexpectedly reboot and XenServer crashes or freezes... 30 6.20. With no NVIDIA driver installed, XenServer misidentifies Tesla M10 cards...31 6.21. GNOME Display Manager (GDM) fails to start on Red Hat Enterprise Linux 7.2 and CentOS 7.0...31 6.22. Video goes blank when run in loop in Windows Media Player... 32 6.23. Local VGA console is momentarily unblanked when XenDesktop changes resolution of the VM desktop... 32 6.24. VM bugchecks on shutdown/restart when XenDesktop is installed and NVIDIA driver is uninstalled or upgraded... 33 6.25. Application frame rate may drop when running XenDesktop at 2560 1600 resolution... 34 6.26. Windows VM BSOD... 34 6.27. Windows VM BSOD when upgrading NVIDIA drivers over a XenDesktop session...35 6.28. XenCenter does not allow vgpus to be selected as a GPU type for Linux VMs...36 6.29. If X server is killed on a RHEL7 VM running vgpu, XenCenter console may not automatically switch to text console... 36 6.30. Multiple WebGL tabs in Microsoft Internet Explorer may trigger TDR on Windows VMs... 37 6.31. XenDesktop shows only a black screen when connected to a vgpu VM... 37 RN-06927-001 _v6.0 Revision 02 iii

RN-06927-001 _v6.0 Revision 02 iv

Chapter 1. RELEASE NOTES These Release Notes summarize current status, information on validated platforms, and known issues with NVIDIA vgpu software and associated hardware on Citrix XenServer. The releases in this release family of NVIDIA vgpu software include the software listed in the following table: Software 6.0 NVIDIA Virtual GPU Manager for the Citrix XenServer releases listed in Hypervisor Software Releases 390.42 NVIDIA Windows driver 391.03 NVIDIA Linux driver version 390.42 Caution If you install the wrong package for the version of Citrix XenServer you are using, NVIDIA Virtual GPU Manager will fail to load. The vgpu Manager and guest VM drivers must be installed together. Older VM drivers will not function correctly with this release of vgpu Manager. Similarly, older releases of vgpu Manager will not function correctly with this release of the guest VM drivers. See VM running older NVIDIA vgpu drivers fails to initialize vgpu when booted. 1.1. Updates in Release 6.0 New Features in Release 6.0 XenMotion support for VMs that are configured with vgpu New -2B vgpu type for each supported GPU vgpu support for NVML accounting functions vgpu support for nvidia-smi accounting modes RN-06927-001 _v6.0 Revision 02 1

Release Notes Change of default scheduler to best effort scheduler for GPUs based on the NVIDIA Pascal architecture Change of maximum resolution for unlicensed GPUs based on the Pascal architecture to 1280 1024 Plain-text logging on Windows of significant licensing events New setting EnableLogging for disabling or enabling logging of significant licensing events Miscellaneous bug fixes Hardware and Software Support Introduced in Release 6.0 Support for GPUs based on the NVIDIA Volta architecture Support for Citrix XenServer 7.4 Support for Windows 10 Fall Creators Update (1709) as a guest OS Feature Support Withdrawn in Release 6.0 Citrix XenServer 6.5 SP1 is no longer supported. RN-06927-001 _v6.0 Revision 02 2

Chapter 2. VALIDATED PLATFORMS This release of NVIDIA vgpu software provides support for several NVIDIA GPUs on validated server hardware platforms, Citrix XenServer hypervisor software versions, and guest operating systems. 2.1. Supported NVIDIA GPUs and Validated Server Platforms This release of NVIDIA vgpu software provides support for the following NVIDIA GPUs on Citrix XenServer, running on validated server hardware platforms: Tesla M6 Tesla M10 Tesla M60 Tesla P4 Tesla P6 Tesla P40 Tesla P100 PCIe 16 GB (XenMotion with vgpu is not supported.) Tesla P100 SXM2 16 GB (XenMotion with vgpu is not supported.) Tesla P100 PCIe 12GB (XenMotion with vgpu is not supported.) Tesla V100 SXM2 (XenMotion with vgpu is not supported.) Tesla V100 PCIe (XenMotion with vgpu is not supported.) Tesla V100 FHHL (XenMotion with vgpu is not supported.) For a list of validated server platforms, refer to NVIDIA GRID Certified Servers. Tesla M60 and M6 GPUs support compute mode and graphics mode. NVIDIA vgpu requires GPUs that support both modes to operate in graphics mode. Recent Tesla M60 GPUs and M6 GPUs are supplied in graphics mode. However, your GPU might be in compute mode if it is an older Tesla M60 GPU or M6 GPU, or if its mode has previously been changed. RN-06927-001 _v6.0 Revision 02 3

Validated Platforms To configure the mode of Tesla M60 and M6 GPUs, use the gpumodeswitch tool provided with NVIDIA vgpu software releases. 2.2. Hypervisor Software Releases This release supports only the hypervisor software releases listed in the table. If a specific release, even an update release, is not listed, it s not supported. Software Release Supported Notes Citrix XenServer 7.4 RTM build is supported. All NVIDIA GPUs that support NVIDIA vgpu software are supported. This release supports XenMotion with vgpu on suitable GPUs as listed in Supported NVIDIA GPUs and Validated Server Platforms. Citrix XenServer 7.3 RTM build is supported. All NVIDIA GPUs that support NVIDIA vgpu software are supported. XenMotion with vgpu is not supported. Citrix XenServer 7.2 RTM build is supported. All NVIDIA GPUs that support NVIDIA vgpu software are supported. XenMotion with vgpu is not supported. Citrix XenServer 7.1 RTM build is supported. All NVIDIA GPUs that support NVIDIA vgpu software are supported. XenMotion with vgpu is not supported. Citrix XenServer 7.0 RTM build 125380 is supported. All NVIDIA GPUs that support NVIDIA vgpu software are supported. XenMotion with vgpu is not supported. Citrix XenDesktop Version 7.15, 7.6, 7.8, and 7.11 in HDX 3D Pro mode. RN-06927-001 _v6.0 Revision 02 4

Validated Platforms 2.3. Guest OS Support NVIDIA vgpu software supports several Windows releases and Linux distributions as a guest OS. The supported guest operating systems depend on the hypervisor software version. Use only a guest OS release that is listed as supported by NVIDIA vgpu software with your virtualization software. To be listed as supported, a guest OS release must be supported not only by NVIDIA vgpu software, but also by your virtualization software. NVIDIA cannot support guest OS releases that your virtualization software does not support. In pass-through mode, GPUs based on the Pascal architecture support only 64-bit guest operating systems. No 32-bit guest operating systems are supported in passthrough mode for these GPUs. 2.3.1. Windows Guest OS Support NVIDIA vgpu software supports only the Windows releases listed in in the table as a guest OS on Citrix XenServer. The releases of Citrix XenServer for which a Windows release is supported depend on whether NVIDIA vgpu or pass-through GPU is used. If a specific release, even an update release, is not listed, it s not supported. XenMotion with vgpu is supported on supported Windows guest OS releases. Guest OS NVIDIA vgpu - Citrix XenServer Releases Pass-Through GPU - Citrix XenServer Releases Windows Server 2016 1607, 1709 Windows Server 2012 R2 Windows Server 2008 R2 Supported only on GPUs based on the Maxwell architecture Windows 10 RTM (1507), November Update (1511), Anniversary Update (1607), Creators Update (1703), Fall Creators Update (1709) (64-bit) Windows 10 RTM (1507), November Update (1511), Anniversary Update (1607), Creators Update (1703), Fall Creators Update (1709) (32-bit) Supported only on GPUs based on the Maxwell and Pascal architectures Supported only on GPUs based on the Maxwell architecture Windows 8.1 Update (64-bit) Windows 8.1 Update (32-bit) RN-06927-001 _v6.0 Revision 02 5

Validated Platforms NVIDIA vgpu - Citrix XenServer Releases Pass-Through GPU - Citrix XenServer Releases Supported only on GPUs based on the Maxwell and Pascal architectures Supported only on GPUs based on the Maxwell architecture Windows 8.1 (64-bit) - Windows 8.1 (32-bit) - Guest OS Supported only on GPUs based on the Maxwell and Pascal architectures Windows 8 (64-bit) - Windows 8 (32-bit) - Supported only on GPUs based on the Maxwell and Pascal architectures Windows 7 (64-bit) Supported only on GPUs based on the Maxwell architecture Windows 7 (32-bit) Supported only on GPUs based on the Maxwell and Pascal architectures Supported only on GPUs based on the Maxwell architecture 2.3.2. Linux Guest OS Support NVIDIA vgpu software supports only the Linux distributions listed in the table as a guest OS on Citrix XenServer. The releases of Citrix XenServer for which a Linux release is supported depend on whether NVIDIA vgpu or pass-through GPU is used. If a specific release, even an update release, is not listed, it s not supported. XenMotion with vgpu is not supported on any Linux guest OS release. Guest OS NVIDIA vgpu - Citrix XenServer Releases Pass-Through GPU - Citrix XenServer Releases Red Hat Enterprise Linux 7.0-7.4 CentOS 7.0-7.4 Ubuntu 16.04 LTS Ubuntu 14.04 LTS RN-06927-001 _v6.0 Revision 02 6

Validated Platforms 2.4. XenMotion with vgpu Support XenMotion with vgpu is supported only on a subset of supported GPUs, Citrix XenServer Hypervisor releases, and guest operating systems. Supported GPUs: Tesla M6 Tesla M10 Tesla M60 Tesla P4 Tesla P6 Tesla P40 Supported Citrix XenServer Hypervisor releases: 7.4 only. Supported guest OS releases: Windows only. XenMotion with vgpu is not supported on Linux. RN-06927-001 _v6.0 Revision 02 7

Chapter 3. KNOWN PRODUCT LIMITATIONS Known product limitations for this release of NVIDIA vgpu software are described in the following sections. 3.1. vgpu profiles with 512 Mbytes or less of frame buffer support only 1 virtual display head on Windows 10 To reduce the possibility of memory exhaustion, vgpu profiles with 512 Mbytes or less of frame buffer support only 1 virtual display head on a Windows 10 guest OS. The following vgpu profiles have 512 Mbytes or less of frame buffer: Tesla M6-0B, M6-0Q Tesla M10-0B, M10-0Q Tesla M60-0B, M60-0Q Use a profile that supports more than 1 virtual display head and has at least 1 Gbyte of frame buffer. RN-06927-001 _v6.0 Revision 02 8

Known Product Limitations 3.2. NVENC requires at least 1 Gbyte of frame buffer Using the frame buffer for the NVIDIA hardware-based H.264/HEVC video encoder (NVENC) may cause memory exhaustion with vgpu profiles that have 512 Mbytes or less of frame buffer. To reduce the possibility of memory exhaustion, NVENC is disabled on profiles that have 512 Mbytes or less of frame buffer. Application GPU acceleration remains fully supported and available for all profiles, including profiles with 512 MBytes or less of frame buffer. NVENC support from both Citrix and VMware is a recent feature and, if you are using an older version, you should experience no change in functionality. The following vgpu profiles have 512 Mbytes or less of frame buffer: Tesla M6-0B, M6-0Q Tesla M10-0B, M10-0Q Tesla M60-0B, M60-0Q If you require NVENC to be enabled, use a profile that has at least 1 Gbyte of frame buffer. 3.3. VM running older NVIDIA vgpu drivers fails to initialize vgpu when booted A VM running a version of the NVIDIA guest VM drivers from a previous main release branch, for example release 4.4, will fail to initialize vgpu when booted on a Citrix XenServer platform running the current release of Virtual GPU Manager. In this scenario, the VM boots in standard VGA mode with reduced resolution and color depth. The NVIDIA virtual GPU is present in Windows Device Manager but displays a warning sign, and the following device status: Windows has stopped this device because it has reported problems. (Code 43) Depending on the versions of drivers in use, the Citrix XenServer VM s /var/log/ messages log file reports one of the following errors: An error message: RN-06927-001 _v6.0 Revision 02 9

Known Product Limitations vmiop_log: error: Unable to fetch Guest NVIDIA driver information A version mismatch between guest and host drivers: vmiop_log: error: Guest VGX version(1.1) and Host VGX version(1.2) do not match A signature mismatch: vmiop_log: error: VGPU message signature mismatch. Resolution Install the current NVIDIA guest VM driver in the VM. 3.4. Virtual GPU fails to start if ECC is enabled Tesla M60, Tesla M6, and GPUs based on the Pascal GPU architecture, for example Tesla P100 or Tesla P4, support error correcting code (ECC) memory for improved data integrity. Tesla M60 and M6 GPUs in graphics mode are supplied with ECC memory disabled by default, but it may subsequently be enabled using nvidia-smi. GPUs based on the Pascal GPU architecture are supplied with ECC memory enabled. However, NVIDIA vgpu does not support ECC memory. If ECC memory is enabled, NVIDIA vgpu fails to start. Citrix XenCenter displays the following error message: An emulator required to run this VM failed to start The following error is logged in the Citrix XenServer host s /var/log/messages log file: vmiop_log: error: Initialization: VGX not supported with ECC Enabled. Resolution Ensure that ECC is disabled on all GPUs. Before you begin, ensure that NVIDIA Virtual GPU Manager is installed on your hypervisor. 1. Use nvidia-smi to list the status of all GPUs, and check for ECC noted as enabled on GPUs. # nvidia-smi -q ==============NVSMI LOG============== Timestamp Driver Version : Tue Dec 19 18:36:45 2017 : 384.99 RN-06927-001 _v6.0 Revision 02 10

Known Product Limitations Attached GPUs GPU 0000:02:00.0 : 1 [...] Ecc Mode Current Pending 2. : Enabled : Enabled [...] Change the ECC status to off on each GPU for which ECC is enabled. If you want to change the ECC status to off for all GPUs on your host machine, run this command: # nvidia-smi -e 0 If you want to change the ECC status to off for a specific GPU, run this command: # nvidia-smi -i id -e 0 id is the index of the GPU as reported by nvidia-smi. This example disables ECC for the GPU with index 0000:02:00.0. 3. 4. # nvidia-smi -i 0000:02:00.0 -e 0 Reboot the host. # shutdown r now Confirm that ECC is now disabled for the GPU. # nvidia-smi -q ==============NVSMI LOG============== Timestamp Driver Version : Tue Dec 19 18:37:53 2017 : 384.99 Attached GPUs GPU 0000:02:00.0 [...] : 1 Ecc Mode Current Pending : Disabled : Disabled [...] If you later need to enable ECC on your GPUs, run one of the following commands: If you want to change the ECC status to on for all GPUs on your host machine, run this command: # nvidia-smi -e 1 If you want to change the ECC status to on for a specific GPU, run this command: # nvidia-smi -i id -e 1 id is the index of the GPU as reported by nvidia-smi. RN-06927-001 _v6.0 Revision 02 11

Known Product Limitations This example enables ECC for the GPU with index 0000:02:00.0. # nvidia-smi -i 0000:02:00.0 -e 1 After changing the ECC status to on, reboot the host. 3.5. Single vgpu benchmark scores are lower than pass-through GPU A single vgpu configured on a physical GPU produces lower benchmark scores than the physical GPU run in pass-through mode. Aside from performance differences that may be attributed to a vgpu s smaller framebuffer size, vgpu incorporates a performance balancing feature known as Frame Rate Limiter (FRL), which is enabled on all vgpus. FRL is used to ensure balanced performance across multiple vgpus that are resident on the same physical GPU. The FRL setting is designed to give good interactive remote graphics experience but may reduce scores in benchmarks that depend on measuring frame rendering rates, as compared to the same benchmarks running on a pass-through GPU. Resolution FRL is controlled by an internal vgpu setting. NVIDIA does not validate vgpu with FRL disabled, but for validation of benchmark performance, FRL can be temporarily disabled by specifying frame_rate_limiter=0 in the VM s platform:vgpu_extra_args parameter: [root@xenserver ~]# xe vm-param-set uuid=e71afda4-53f4-3a1b-6c92-a364a7f619c2 platform:vgpu_extra_args="frame_rate_limiter=0" [root@xenserver ~]# The setting takes effect the next time the VM is started or rebooted. With this setting in place, the VM s vgpu will run without any frame rate limit. The FRL can be reverted back to its default setting in one of the following ways: Removing the vgpu_extra_args key from the platform parameter Removing frame_rate_limiter=0 from the vgpu_extra_args key Setting frame_rate_limiter=1. For example: [root@xenserver ~]# xe vm-param-set uuid=e71afda4-53f4-3a1b-6c92a364a7f619c2 platform:vgpu_extra_args="frame_rate_limiter=1" [root@xenserver ~]# RN-06927-001 _v6.0 Revision 02 12

Known Product Limitations 3.6. nvidia-smi fails to operate when all GPUs are assigned to GPU pass-through mode If all GPUs in the platform are assigned to VMs in pass-through mode, nvidia-smi will return an error: [root@xenserver-vgx-test ~]# nvidia-smi Failed to initialize NVML: Unknown Error This is because GPUs operating in pass-through mode are not visible to nvidia-smi and the NVIDIA kernel driver operating in the Citrix XenServer dom0. To confirm that all GPUs are operating in pass-through mode, use XenCenter s GPU tab to review current GPU assignment: RN-06927-001 _v6.0 Revision 02 13

Known Product Limitations Resolution N/A 3.7. Windows Aero is disabled on XenDesktop session using 3 or 4 monitors in 2560 1600 resolution Windows Aero may be disabled when XenDesktop is connected to a VM with a vgpu or passthrough GPU, with 3 or 4 monitors at 2560 1600 resolution. This limitation is a limitation of Windows 7. For details, see the Microsoft knowledge base article Desktop background disappears with very large extended desktop on Windows 7. 3.8. VMs configured with large memory fail to initialize vgpu when booted When starting multiple VMs configured with large amounts of RAM (typically more than 32GB per VM), a VM may fail to initialize vgpu. In this scenario, the VM boots in standard VGA mode with reduced resolution and color depth. The NVIDIA vgpu software GPU is present in Windows Device Manager but displays a warning sign, and the following device status: Windows has stopped this device because it has reported problems. (Code 43) The Citrix XenServer VM s /var/log/messages log file contains these error messages: vmiop_log: vmiop_log: vmiop_log:... vmiop_log:... vmiop_log: vmiop_log: vmiop_log:... vmiop_log: error: NVOS status 0x29 error: Assertion Failed at 0x7620fd4b:179 error: 8 frames returned by backtrace error: VGPU message 12 failed, result code: 0x29 error: NVOS status 0x8 error: Assertion Failed at 0x7620c8df:280 error: 8 frames returned by backtrace error: VGPU message 26 failed, result code: 0x8 RN-06927-001 _v6.0 Revision 02 14

Known Product Limitations Resolution vgpu reserves a portion of the VM s framebuffer for use in GPU mapping of VM system memory. The reservation is sufficient to support up to 32GB of system memory, and may be increased to accommodate up to 64GB by specifying enable_large_sys_mem=1 in the VM s platform:vgpu_extra_args parameter: [root@xenserver ~]# xe vm-param-set uuid=e71afda4-53f4-3a1b-6c92-a364a7f619c2 platform:vgpu_extra_args="enable_large_sys_mem=1" The setting takes effect the next time the VM is started or rebooted. With this setting in place, less GPU FB is available to applications running in the VM. To accommodate system memory larger than 64GB, the reservation can be further increased by specifying extra_fb_reservation in the VM s platform:vgpu_extra_args parameter, and setting its value to the desired reservation size in megabytes. The default value of 64M is sufficient to support 64GB of RAM. We recommend adding 2M of reservation for each additional 1GB of system memory. For example, to support 96GB of RAM, set extra_fb_reservation to 128: platform:vgpu_extra_args="enable_large_sys_mem=1, extra_fb_reservation=128" The reservation can be reverted back to its default setting in one of the following ways: Removing the vgpu_extra_args key from the platform parameter Removing enable_large_sys_mem from the vgpu_extra_args key Setting enable_large_sys_mem=0 3.9. vgpu host driver RPM upgrade fails Upgrading vgpu host driver RPM fails with an error message about failed dependencies on the console. [root@xenserver ~]# rpm U NVIDIA-vGPU-xenserver-6.5-352.46.x86_64.rpm error: Failed dependencies: NVIDIA-vgx-xenserver conflicts with NVIDIA-vGPUxenserver-6.5-352.46.x86_64 [root@xenserver ~]# Resolution Uninstall the older vgpu RPM before installing the latest driver. Use the following command to uninstall the older vgpu RPM: [root@xenserver ~]# rpm e NVIDIA-vgx-xenserver RN-06927-001 _v6.0 Revision 02 15

Chapter 4. RESOLVED ISSUES Issues Resolved in Release 6.0 No resolved issues are reported in this release for Citrix XenServer. RN-06927-001 _v6.0 Revision 02 16

Chapter 5. NVIDIA SOFTWARE SECURITY UPDATES For more information about NVIDIA s vulnerability management, visit the NVIDIA Product Security page. NVIDIA Software Security Updates in Release No NVIDIA software security updates are reported in this release for Citrix XenServer. RN-06927-001 _v6.0 Revision 02 17

Chapter 6. KNOWN ISSUES 6.1. Migrating a VM Configured with vgpu Fails Migrating a VM configured with vgpu fails if vgpu migration is not configured in NVIDIA vgpu software. Resolution Ensure that vgpu migration is configured in NVIDIA vgpu software to enable support for XenMotion. For instructions, see Virtual GPU Software User Guide. 6.2. vgpu guest VM driver not properly loaded on servers with more than 512 GB of system memory Support for vgpu is limited to servers with a maximum of 512 GB of system memory. On servers with more than 512 GB of system memory, the guest VM driver is not properly loaded. Device Manager marks the vgpu with a yellow exclamation point. Resolution Disable PV IOMMU to limit the amount of system memory on the server to a maximum of 512 GB. [root@xenserver ~]# /opt/xensource/libexec/xen-cmdline --set-xen iommu=dom0passhthrough RN-06927-001 _v6.0 Revision 02 18

Known Issues Not an NVIDIA bug 1799582 6.3. Issues in remote desktop sessions if a license is acquired after a session is started A VM might acquire a license for NVIDIA vgpu software after a remote desktop session has connected to the VM. In this situation, some licensed features and capabilities are not available to a properly licensed vgpu or pass-through GPU in the session. For example, the updated maximum resolution supported is not available. Before attempting this workaround, confirm that the VM has obtained the correct license for NVIDIA vgpu software. 1. 2. After installing the guest VM driver package and configuring required license settings on the VM (or on the master image used for VM deployment), add the following registry setting: [HKEY_LOCAL_MACHINE\SOFTWARE\NVIDIA Corporation\Global\GridLicensing] "IgnoreSP"=dword:00000001 Restart the VM. Open 200391532 6.4. The license expires prematurely in Linux guest VMs In Linux guest VMs, the license expires before the default borrow period has elapsed. In normal operation, the license is renewed periodically at an interval that depends on the RN-06927-001 _v6.0 Revision 02 19

Known Issues license borrow period. As a result, a failure to renew the license may cause the license to expire before the default borrow period has elapsed. To reduce the possibility of license-renewal failures caused by transient network issues, increase the license borrow period to a value of about 7 days. Open 200376678 6.5. Multiple display heads are not detected by Ubuntu 14.04 guest VMs After an Ubuntu 14.04 guest VM has acquired a license, multiple display heads connected to the VM are not detected. Version Ubuntu 14.04 To see all the connected display heads after the VM has acquired a license, open the Displays settings window and click Detect displays. Open 200334648 RN-06927-001 _v6.0 Revision 02 20

Known Issues 6.6. CILP is not working in Windows VMs CILP is not working in Windows VMs. If a CUDA application is running in one VM and a graphics application is running in another VM, the following errors occur: The CUDA application times out. A TDR is triggered in the VM that is running graphics application, which may cause flickering or an application crash. CILP isn't expected to work on Windows until Windows 10 RS3. Version Windows 10 RS2 Creators Update Open 200333574 6.7. Luxmark causes a segmentation fault on an unlicensed Linux client If the Luxmark application is run on a Linux guest VM configured with NVIDIA vgpu that is booted without acquiring a license, a segmentation fault occurs and the application core dumps. The fault occurs when the application cannot allocate a CUDA object on NVIDIA vgpus where CUDA is disabled. On NVIDIA vgpus that can support CUDA, CUDA is disabled in unlicensed mode. Not an NVIDIA bug. 200330956 RN-06927-001 _v6.0 Revision 02 21

Known Issues 6.8. On GPUs based on the Pascal architecture, Ubuntu 16.04 VMs run slowly after acquiring a license On GPUs based on the Pascal architecture, Ubuntu VMs to which an NVIDIA vgpu or pass-through GPU is assigned run slowly after acquiring a license. Ubuntu VMs that have not been assigned an NVIDIA vgpuor pass-through GPU run noticeably faster. After the VM has acquired a license, restart the lightdm service. Open. 200359618 6.9. Resolution is not updated after a VM acquires a license and is restarted In a Red Enterprise Linux 7.3 guest VM, an increase in resolution from 1024 768 to 2560 1600 is not applied after a license is acquired and the gridd service is restarted. This issue occurs if the multimonitor parameter is added to the xorg.conf file. Version Red Enterprise Linux 7.3 Open 200275925 RN-06927-001 _v6.0 Revision 02 22

Known Issues 6.10. On Tesla P40, P6, and P4 GPUs, the default ECC setting prevents NVIDIA vgpu from starting On Tesla P40, Tesla P6, and Tesla P4 GPUs, the default error-correcting code (ECC) memory setting prevents NVIDIA vgpu from starting. By default, ECC memory is enabled on these GPUs, but NVIDIA vgpu does not support ECC memory. Before running NVIDIA vgpu, disable ECC memory as explained in Virtual GPU Software User Guide. Closed. 200269717 6.11. NVIDIA vgpu encoder and process utilization counters don't work with Windows Performance Counters GPU encoder and process utilization counter groups are listed in Windows Performance Counters, but no instances of the counters are available. The counters are disabled by default and must be enabled. Enable the counters by running the following sequence of commands from a command shell: wmic /namespace:nv path System call enableprocessutilizationperfcounter wmic /namespace:nv path System call enableencodersessionsperfcounter If you need to disable the counters, run the following sequence of commands from a command shell: RN-06927-001 _v6.0 Revision 02 23

Known Issues wmic /namespace:nv path System call disableprocessutilizationperfcounter wmic /namespace:nv path System call disableencodersessionsperfcounter Open 1971698 6.12. A segmentation fault in DBus code causes nvidia-gridd to exit on Red Hat Enterprise Linux and CentOS On Red Hat Enterprise Linux 6.8 and 6.9, and CentOS 6.8 and 6.9, a segmentation fault in DBus code causes the nvidia-gridd service to exit. The nvidia-gridd service uses DBus for communication with NVIDIA X Server Settings to display licensing information through the Manage License page. Disabling the GUI for licensing resolves this issue. To prevent this issue, the GUI for licensing is disabled by default. You might encounter this issue if you have enabled the GUI for licensing and are using Red Hat Enterprise Linux 6.8 or 6.9, or CentOS 6.8 and 6.9. Version Red Hat Enterprise Linux 6.8 and 6.9 CentOS 6.8 and 6.9 Open 200358191 200319854 1895945 RN-06927-001 _v6.0 Revision 02 24

Known Issues 6.13. No Manage License option available in NVIDIA X Server Settings by default By default, the Manage License option is not available in NVIDIA X Server Settings. This option is missing because the GUI for licensing on Linux is disabled by default to work around the issue that is described in A segmentation fault in DBus code causes nvidia-gridd to exit on Red Hat Enterprise Linux and CentOS. This workaround requires sudo privileges. Do not use this workaround with Red Hat Enterprise Linux 6.8 and 6.9 or CentOS 6.8 and 6.9. To prevent a segmentation fault in DBus code from causing the nvidiagridd service from exiting, the GUI for licensing must be disabled with these OS versions. If NVIDIA X Server Settings is running, shut it down. 2. If the /etc/nvidia/gridd.conf file does not already exist, create it by copying the supplied template file /etc/nvidia/gridd.conf.template. 3. As root, edit the /etc/nvidia/gridd.conf file to set the EnableUI option to TRUE. 4. Start the nvidia-gridd service. 1. # sudo service nvidia-gridd start When NVIDIA X Server Settings is restarted, the Manage License option is now available. Open 6.14. The nvidia-gridd service fails because the required configuration is not provided The nvidia-gridd service exits with an error because the required configuration is not provided. RN-06927-001 _v6.0 Revision 02 25

Known Issues The known issue described in A segmentation fault in DBus code causes nvidia-gridd to exit on Red Hat Enterprise Linux and CentOS causes the NVIDIA X Server Settings page for managing licensing settings through a GUI to be disabled by default. As a result, if the required license configuration is not provided through the configuration file, the service exits with an error. Details of the error can be obtained by checking the status of the nvidia-gridd service. # service nvidia-gridd status nvidia-gridd.service - NVIDIA Grid Daemon Loaded: loaded (/usr/lib/systemd/system/nvidia-gridd.service; enabled; vendor preset: disabled) Active: failed (Result: exit-code) since Wed 2017-11-01 19:25:07 IST; 27s ago Process: 11990 ExecStopPost=/bin/rm -rf /var/run/nvidia-gridd (code=exited, status=0/success) Process: 11905 ExecStart=/usr/bin/nvidia-gridd (code=exited, status=0/success) Main PID: 11906 (code=exited, status=1/failure) Nov 01 19:24:35 localhost.localdomain systemd[1]: Starting NVIDIA Grid Daemon... Nov 01 19:24:35 localhost.localdomain nvidia-gridd[11906]: Started (11906) Nov 01 19:24:35 localhost.localdomain systemd[1]: Started NVIDIA Grid Daemon. Nov 01 19:24:36 localhost.localdomain nvidia-gridd[11906]: Failed to open config file : /etc/nvidia/gridd.conf error :No such file or directory Nov 01 19:25:07 localhost.localdomain nvidia-gridd[11906]: Service provider detection complete. Nov 01 19:25:07 localhost.localdomain nvidia-gridd[11906]: Shutdown (11906) Nov 01 19:25:07 localhost.localdomain systemd[1]: nvidia-gridd.service: main process exited, code=exited, status=1/failure Nov 01 19:25:07 localhost.localdomain systemd[1]: Unit nvidia-gridd.service entered failed state. Nov 01 19:25:07 localhost.localdomain systemd[1]: nvidia-gridd.service failed. Use a configuration file to license NVIDIA vgpu software on Linux as explained in Virtual GPU Client Licensing User Guide. Open 200359469 6.15. The Apply button is disabled after change to unlicensed mode After the mode is changed from licensed Quadro Virtual Datacenter Workstation Edition mode to Unlicensed Tesla mode, the Apply button on the Manage GRID RN-06927-001 _v6.0 Revision 02 26

Known Issues License page is disabled. As a result, NVIDIA X Server Settings cannot be used to switch to Tesla (Unlicensed) mode on a licensed system. 1. 2. 3. 4. 5. Start NVIDIA X Server Settings by using the method for launching applications provided by your Linux distribution. In the NVIDIA X Server Settings window that opens, click Manage GRID License. Clear the Primary Server field. Select the Tesla (unlicensed) option. Click Apply. Open 200359624 6.16. Licenses remain checked out when VMs are forcibly powered off NVIDIA vgpu software licenses remain checked out on the license server when nonpersistent VMs are forcibly powered off. The NVIDIA service running in a VM returns checked out licenses when the VM is shut down. In environments where non-persistent licensed VMs are not cleanly shut down, licenses on the license server can become exhausted. For example, this issue can occur in automated test environments where VMs are frequently changing and are not guaranteed to be cleanly shut down. The licenses from such VMs remain checked out against their MAC address for seven days before they time out and become available to other VMs. Resolution If VMs are routinely being powered off without clean shutdown in your environment, you can avoid this issue by shortening the license borrow period. To shorten the license borrow period, set the LicenseInterval configuration setting in your VM image. For details, refer to Virtual GPU Client Licensing User Guide. RN-06927-001 _v6.0 Revision 02 27

Known Issues Closed 1694975 6.17. Memory exhaustion can occur with vgpu profiles that have 512 Mbytes or less of frame buffer Memory exhaustion can occur with vgpu profiles that have 512 Mbytes or less of frame buffer. This issue typically occurs in the following situations: Full screen 1080p video content is playing in a browser. In this situation, the session hangs and session reconnection fails. Multiple display heads are used with Citrix XenDesktop or VMware Horizon on a Windows 10 guest VM. Higher resolution monitors are used. Applications that are frame-buffer intensive are used. NVENC is in use. To reduce the possibility of memory exhaustion, NVENC is disabled on profiles that have 512 Mbytes or less of frame buffer. When memory exhaustion occurs, the NVIDIA host driver reports Xid error 31 and Xid error 43 in XenServer s /var/log/messages file. The following vgpu profiles have 512 Mbytes or less of frame buffer: Tesla M6-0B, M6-0Q Tesla M10-0B, M10-0Q Tesla M60-0B, M60-0Q The root cause is a known issue associated with changes to the way that recent Microsoft operating systems handle and allow access to overprovisioning messages and errors. If your systems are provisioned with enough frame buffer to support your use cases, you should not encounter these issues. RN-06927-001 _v6.0 Revision 02 28

Known Issues Use an appropriately sized vgpu to ensure that the frame buffer supplied to a VM through the vgpu is adequate for your workloads. Monitor your frame buffer usage. If you are using Windows 10, consider these workarounds and solutions: Use a profile that has 1 Gbyte of frame buffer. Optimize your Windows 10 resource usage. To obtain information about best practices for improved user experience using Windows 10 in virtual environments, complete the NVIDIA GRID vgpu Profile Sizing Guide for Windows 10 download request form. For more information, see also Windows 10 Optimization for XenDesktop on the Citrix blog. Open 200130864 1803861 6.18. VM bug checks after the guest VM driver for Windows 10 RS2 is installed When the VM is rebooted after the guest VM driver for Windows 10 RS2 is installed, the VM bug checks. When Windows boots, it selects one of the standard supported video modes. If Windows is booted directly with a display that is driven by an NVIDIA driver, for example a vgpu on Citrix XenServer, a blue screen crash occurs. This issue occurs when the screen resolution is switched from VGA mode to a resolution that is higher than 1920 1200. Fix Download and install Microsoft Windows Update KB4020102 from the Microsoft Update Catalog. If you have applied the fix, ignore this workaround. RN-06927-001 _v6.0 Revision 02 29

Known Issues Otherwise, you can work around this issue until you are able to apply the fix by not using resolutions higher than 1920 1200. Choose a GPU profile in Citrix XenCenter that does not allow resolutions higher than 1920 1200. 2. Before rebooting the VM, set the display resolution to 1920 1200 or lower. 1. Not an NVIDIA bug 200310861 6.19. On XenServer 7.0, VMs unexpectedly reboot and XenServer crashes or freezes On XenServer 7.0, VMs to which a vgpu is attached unexpectedly reboot and XenServer crashes or freezes. The event log in XenServer s /var/log/crash/xen.log file lists the following errors: A fatal bus error on a component at the slot where the GPU card is installed A fatal error on a component at bus 0, device 2, function 0 This issue occurs when page-modification logging (PML) is enabled on Intel Broadwell CPUs running XenServer 7.0. Citrix is aware of this issue and is working on a permanent fix. Disable page-modification logging (PML) as explained in XenServer 7 host crash while starting multiple virtual machines in the Citrix Support Knowledge Center. Not an NVIDIA bug 1853248 RN-06927-001 _v6.0 Revision 02 30

Known Issues 6.20. With no NVIDIA driver installed, XenServer misidentifies Tesla M10 cards An erroneous entry in the pci.ids database causes Citrix XenServer to identify Tesla M10 cards as GRID M40 when no NVIDIA driver is installed. Version Citrix XenServer 6.5 and 7.0 None Not an NVIDIA bug NVIDIA-420/1792341 6.21. GNOME Display Manager (GDM) fails to start on Red Hat Enterprise Linux 7.2 and CentOS 7.0 GDM fails to start on Red Hat Enterprise Linux 7.2 and CentOS 7.0 with the following error: Oh no! Something has gone wrong! Permanently enable permissive mode for Security Enhanced Linux (SELinux). 1. 2. As root, edit the /etc/selinux/config file to set SELINUX to permissive. SELINUX=permissive Reboot the system. ~]# reboot RN-06927-001 _v6.0 Revision 02 31

Known Issues For more information, see Permissive Mode in Red Hat Enterprise Linux 7 SELinux User's and Administrator's Guide. Not an NVIDIA bug 200167868 6.22. Video goes blank when run in loop in Windows Media Player When connected to a vgpu-enabled VM using Citrix XenDesktop, a video played back in looping mode on Windows Media Player goes blank or freezes after a few iterations. None Not an NVIDIA bug 1306623 6.23. Local VGA console is momentarily unblanked when XenDesktop changes resolution of the VM desktop When XenDesktop establishes a remote connection to a VM using vgpu, the VM s local VGA console display in XenCenter is blanked (assuming the VM local console has not been disabled by setting platform:vgpu_extra_args="disable_vnc=1"). If the XenDesktop session changes resolution of the VM s desktop, the local VGA console momentarily unblanks, allowing a XenCenter user to briefly view the desktop. RN-06927-001 _v6.0 Revision 02 32

Known Issues Disable the VM s local VGA console xe vm-param-set uuid=vm-uuid platform:vgpu_extra_args="disable_vnc=1" Open NVIDIA-145/1375164 6.24. VM bugchecks on shutdown/restart when XenDesktop is installed and NVIDIA driver is uninstalled or upgraded. If the XenDesktop agent is installed in a VM before any NVIDIA GPU driver is installed, the VM will bugcheck (bluescreen) when the NVIDIA driver is subsequently upgraded or uninstalled. The bugcheck code is 0x7E, SYSTEM_THREAD_EXCEPTION_NOT_HANDLED. Use one of the following workarounds: Do a force shutdown of the VM and restart it. Install the NVIDIA driver in guest VMs before installing XenDesktop. Open NVIDIA-295/200018125 RN-06927-001 _v6.0 Revision 02 33

Known Issues 6.25. Application frame rate may drop when running XenDesktop at 2560 1600 resolution. An application s rendering frame rate may drop when running XenDesktop at 2560 1600 resolution, relative to the frame rate obtained at lower resolutions. Fix Using the Windows regedit utility within the VM, open the HKLM\SOFTWARE \Citrix\Graphics registry key and create a new DWORD value, EncodeSpeed, with a value of 2. Reboot the VM. This setting may improve the delivered frame rate at the expense of a reduction in image quality. Open NVIDIA-190/1416336 6.26. Windows VM BSOD Windows VM bugchecks on XenServer when running a large number of vgpu based VMs. XenServer s /var/log/messages file contains these error messages: NVRM: Xid (PCI:0000:08:00): 31, Ch 0000001e, engmask 00000111, intr 10000000 NVRM: Xid (PCI:0000:08:00): 31, Ch 00000016, engmask 00000111, intr 10000000... vmiop_log: error: Assertion Failed at 0xb5b898d8:4184 vmiop_log: error: 8 frames returned by backtrace vmiop_log: error: /usr/lib/libnvidia-vgx.so(_nv000793vgx+0x69d) [0xb5b8064d] vmiop_log: error: /usr/lib/libnvidia-vgx.so(_nv000479vgx+0x118) [0xb5b898d8] vmiop_log: error: /usr/lib/libnvidia-vgx.so(_nv000782vgx+0x59) [0xb5b85f49] vmiop_log: error: /usr/lib/libnvidia-vgx.so(_nv000347vgx+0x3db) [0xb5b932db] vmiop_log: error: /usr/lib/libnvidia-vgx.so [0xb5b78e4a] vmiop_log: error: /usr/lib/xen/bin/vgpu [0x80554be] vmiop_log: error: /lib/libpthread.so.0 [0xb7612912] vmiop_log: error: /lib/libc.so.6(clone+0x5e) [0xb76fc5ee] vmiop_log: error: failed to initialize guest PTE entries vmiop_log: error: failed to fill up guest PTE entries 3 vmiop_log: error: VGPU message 27 failed, result code: 0xff000003 vmiop_log: error: 0xc1d00001, 0xff010000, 0x1a77ba000, 0x0, 0x1, RN-06927-001 _v6.0 Revision 02 34

Known Issues vmiop_log: error: 0x1, 0x1000, 0x10202, 0xc1d00001, 0xff010000, vmiop_log: error: 0xcaf00004, 0x0 vmiop_log: error: Timeout occurred, reset initiated. Version XenServer 6.2 Fix Ensure that you are running the latest OEM firmware for your NVIDIA vgpu software boards. Closed NVIDIA-327/1632120 6.27. Windows VM BSOD when upgrading NVIDIA drivers over a XenDesktop session Windows VM bugchecks when NVIDIA guest drivers are upgraded over a XenDesktop session. If the VM is restarted after the bugcheck, the upgraded driver loads correctly and full functionality is available. Fix Upgrade XenDesktop to 7.6 Feature Pack 3 Closed NVIDIA-370/200130780 RN-06927-001 _v6.0 Revision 02 35

Known Issues 6.28. XenCenter does not allow vgpus to be selected as a GPU type for Linux VMs When creating a new Linux VM or editing the properties of an existing Linux VM, XenCenter does not allow vgpus to be selected as a GPU type. vgpu on Linux VMs is supported as a technical preview on XenServer 6.5, and does include XenCenter integration. Version Affects the XenCenter integration with XenServer 6.5 only. Resolved in the XenCenter integration with XenServer 7.0. Refer to XenServer vgpu Management in Virtual GPU Software User Guide for how to configure vgpu by using the xe CLI. Closed NVIDIA-360 6.29. If X server is killed on a RHEL7 VM running vgpu, XenCenter console may not automatically switch to text console If X server is killed on a RHEL7 VM running vgpu, XenCenter console may display a corrupted image and fail to switchover to text console. The failure to switchover to text console is due to a bug in RHEL7, which causes X server to not start correctly under certain configurations. RN-06927-001 _v6.0 Revision 02 36

Known Issues Use CTRL+ALT+F1, F2, or F3 to switch between Linux terminals. Closed NVIDIA-350/200123378 6.30. Multiple WebGL tabs in Microsoft Internet Explorer may trigger TDR on Windows VMs Running intensive WebGL applications in multiple IE tabs may trigger a TDR on Windows VMs. Disable hardware acceleration in IE. To enable software rendering in IE, refer to the Microsoft knowledge base article How to enable or disable software rendering in Internet Explorer. Open 200148377 6.31. XenDesktop shows only a black screen when connected to a vgpu VM XenDesktop sometimes displays only a black screen when it is connected to an NVIDIA vgpu VM. The probable cause is that the display that is connected to the NVIDIA vgpu is entering a lower power state. RN-06927-001 _v6.0 Revision 02 37

Known Issues Fix Disable all display-related power management settings. For detailed instructions, visit Microsoft power plans frequently asked questions and from the list, select your OS version. Not an NVIDIA bug 1719877 RN-06927-001 _v6.0 Revision 02 38

Notice ALL NVIDIA DESIGN SPECIFICATIONS, REFERENCE BOARDS, FILES, DRAWINGS, DIAGNOSTICS, LISTS, AND OTHER DOCUMENTS (TOGETHER AND SEPARATELY, "MATERIALS") ARE BEING PROVIDED "AS IS." NVIDIA MAKES NO WARRANTIES, EXPRESSED, IMPLIED, STATUTORY, OR OTHERWISE WITH RESPECT TO THE MATERIALS, AND EXPRESSLY DISCLAIMS ALL IMPLIED WARRANTIES OF NONINFRINGEMENT, MERCHANTABILITY, AND FITNESS FOR A PARTICULAR PURPOSE. Information furnished is believed to be accurate and reliable. However, NVIDIA Corporation assumes no responsibility for the consequences of use of such information or for any infringement of patents or other rights of third parties that may result from its use. No license is granted by implication of otherwise under any patent rights of NVIDIA Corporation. Specifications mentioned in this publication are subject to change without notice. This publication supersedes and replaces all other information previously supplied. NVIDIA Corporation products are not authorized as critical components in life support devices or systems without express written approval of NVIDIA Corporation. HDMI HDMI, the HDMI logo, and High-Definition Multimedia Interface are trademarks or registered trademarks of HDMI Licensing LLC. OpenCL OpenCL is a trademark of Apple Inc. used under license to the Khronos Group Inc. Trademarks NVIDIA, the NVIDIA logo, NVIDIA GRID, vgpu, Pascal, Quadro, and Tesla are trademarks or registered trademarks of NVIDIA Corporation in the U.S. and other countries. Other company and product names may be trademarks of the respective companies with which they are associated. Copyright 2013-2018 NVIDIA Corporation. All rights reserved.